Flach, Smith, Dittman, & Stanard's (2003) hypothesis that a linear weighting of optical angle and expansion rate are used as a criterion for braking is evaluated in a simple driving task. The task required participants to drive from an initial stationary position to a full stop at a comfortable distance in front of a stationary obstacle. The simulated vehicle was controlled with an accelerator and a brake. Acceleration and deceleration were proportional to displacement of these controls. Results showed that people's critical actions (releasing the accelerator and initiating the brake) corresponded with a Tau criterion (constant ratio of angle and expansion rate). Time histories suggest that people were adopting non-proportional, discrete control strategies that approached time optimal, bang-bang control.
Available at: http://works.bepress.com/john_flach/48/